YouTube Transcript Pipeline

ReviewAudited by ClawScan on May 10, 2026.

Overview

The skill is coherent for transcript production, but users should notice that it relies on third-party transcription/translation services and can persist workflow notes.

This skill appears aligned with its transcript-workflow purpose. Before using it, confirm that the YouTube content and interview transcript may be sent to Deepgram and Google Translate, use a scoped Deepgram key, avoid confidential content unless approved, and clarify what durable memory logs will retain.

Findings (4)

Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.

What this means

Providing the key may consume paid quota or expose transcription-account access if handled carelessly.

Why it was flagged

The skill asks for a Deepgram credential even though the registry metadata declares no primary credential. This is expected for Deepgram transcription, but it is still account authority.

Skill content
- Deepgram API key (Nova-3 for best results)
Recommendation

Use a scoped Deepgram key, provide it through a secure secret mechanism when possible, and revoke or rotate it if it may have been exposed.

What this means

Private interview text could leave the local workspace and may be processed or logged by the translation provider or intermediaries.

Why it was flagged

The translation helper sends transcript text to Google Translate as a request parameter. This is purpose-aligned for translation, but it moves transcript content to an external provider.

Skill content
'q': text, ... url = 'https://translate.googleapis.com/translate_a/single?ie=UTF-8&oe=UTF-8&hl=zh-CN&' + qs.replace('+','%20')
Recommendation

Get user approval before translating sensitive transcripts, redact confidential content, or use an approved translation service with suitable data-handling terms.

What this means

Transcript corrections, workflow details, or other notes may persist beyond the current task and could influence later work.

Why it was flagged

The skill instructs the agent to write durable memory logs. This is disclosed, but the artifacts do not define exactly what is logged or retained.

Skill content
Append changes in memory logs for durable state (per workspace conventions).
Recommendation

Keep durable logs minimal, avoid storing sensitive transcript content, and clear or disable memory logging for confidential projects.

What this means

If the agent or user supplies that missing script from elsewhere, it could handle audio files and credentials in ways not covered by this review.

Why it was flagged

The workflow suggests running a helper script that is not included in the provided file manifest, so its behavior and provenance cannot be reviewed here.

Skill content
python3 resilient_dg_transcribe_resume.py \
  --audio-dir /tmp/dg_chunks_work2
Recommendation

Use only reviewed helper scripts, include referenced utilities in the skill package, or require the user to provide a trusted local implementation.